Agreement/Disagreement Classification: Exploiting Unlabeled Data using Contrast Classifiers.
Sangyun HahnRichard E. LadnerMari OstendorfPublished in: HLT-NAACL (2006)
Keyphrases
- unlabeled data
- labeled and unlabeled data
- co training
- class labels
- labeled data
- semi supervised learning
- supervised learning
- training set
- active learning
- supervised learning algorithms
- semi supervised
- training data
- decision boundary
- labeled training data
- text classification
- probabilistic classifiers
- semi supervised classification
- unlabeled samples
- supervised classification
- improve the classification accuracy
- partially labeled data
- classification algorithm
- training examples
- labeled examples
- transductive learning
- unlabeled examples
- learning algorithm
- labeled and unlabeled examples
- training samples
- support vector
- decision trees
- classification models
- supervised classifiers
- text categorization
- machine learning
- naive bayes
- accurate classifiers
- support vector machine
- classification accuracy
- label information
- semi supervised learning algorithms
- learning tasks
- pattern classification
- svm classifier
- instance selection
- select relevant features
- supervised and semi supervised
- class distribution
- unsupervised learning
- data points
- multi view
- feature selection
- positive and unlabeled examples
- feature vectors
- feature set
- machine learning algorithms
- classification systems
- feature space
- positive examples
- feature extraction
- text classifiers
- classifier ensemble